Turn customer reviews into evidence-backed buyer personas
Use Codex and role subagents to turn a CSV of reviews into personas and evidence-backed purchase triggers.
Our suggested starting prompt — not the author's original · untested
You'll need: Codex access and a CSV of customer reviews. Exact subagent prompts and whose reviews were used are not disclosed.
Using the customer-review CSV at [path], extract recurring pain points, motivations, desires, objections and exact customer phrases. For each theme, count how many reviews support it and list example review IDs/quotes. Then spawn five specialist subagents (Customer Researcher, Copywriter, Paid Social Creative Strategist, Organic Social Strategist, Media Buyer). Have each propose buyer personas with 10 evidence-backed 'why now?' purchase triggers. Every trigger must cite at least one review from the CSV and one external customer signal (e.g. a public forum quote) or be dropped. Have the five agents critique each other until they agree on one final document with personas, triggers, objections, required proof and evidence. Do not invent counts — flag anything you could not verify.

